Basecamp Research secures $60 million to develop a 'GPT for biology'
While companies like OpenAI and Anthropic continue to promote the use of everyday language to request responses from artificial intelligence agents, the interest in this form of interaction continues to grow.
A new London startup, Basecamp Research, has successfully raised $60 million to develop an artificial intelligence agent focused on biology and natural biodiversity. This initiative comes at a time when companies like OpenAI and Anthropic have popularized the use of natural language to interact with AI assistants, but Basecamp Research aims to go further, generating new insights that are not possible for humans. Glen Gowers, co-founder and CEO of the company, emphasized in an interview that there is a significant data gap in the training of biological models. He stated that some of the leading pharmaceutical companies in the world are training models without having sufficient information about nature.
To date, Basecamp Research has established over 100 partnerships with organizations in 25 countries to enrich its database with information from primary sources. Approximately 15 of these collaborations are using AI for the development of new products. For example, Procter & Gamble is applying these models to design enzymes that clean stains at low temperatures, while Colorifix is working on more sustainable dye formulations for fabrics.
One of Basecamp Research's most notable achievements is that its BaseFold model outperforms AlphaFold 2, which recently received a Nobel Prize in Chemistry for its creators at DeepMind, in the accurate prediction of complex protein structures and small molecule interactions. The founders' vision for building AI in biology is ambitious; they are developing their models from scratch. Gowers and his partner Oliver Vince, both PhDs in biology, met at the University of Oxford, and the startup derives its name from the time they spent on an ice cap performing DNA sequencing with technology they designed themselves.
Basecamp Research has adapted components of that equipment into smaller units to collect data. In biology, there have been hundreds of books and millions of research pages, but much of this information is outdated, disorganized, and unreliable. To overcome these challenges, the startup is meticulously gathering primary data to build its models from the ground up, with the aim of developing an AI capable of providing better insights in biology than humans can achieve.
Gowers explained that their approach combines global exploration to collect data on natural phenomena such as geysers and volcanoes with an artificial intelligence program aimed at training massive language models to create a system that acts like a "ChatGPT" for nature. The startup has created what could be the "largest computing cluster" dedicated to studying the natural world. Both founders are aware that humanity has captured only about 1% of the available information about biodiversity, making it difficult to formulate the right questions. In summary, Andy Conrad, one of the investors, noted that Basecamp Research's platform has the capability to address questions that the biopharma industry has not yet considered.
The recent Series B funding round, led by the European firm Singular, is complemented by a long-term collaboration with Dr. David R. Liu and the Broad Institute. The funds will be used to further develop the startup, expand its models, and forge partnerships with other biomedical organizations. As for the possibility of a "ChatGPT"-style interface, Gowers clarified that the company is currently more focused on B2B collaborations rather than creating a product aimed at the general public, although they did not rule out that option in the future.
To date, Basecamp Research has raised a total of $85 million, with its latest valuation standing at $71 million, according to PitchBook. The Series B round also included prominent investors such as André Hoffmann, Feike Sijbesma, and Paul Polman, among others.
